Webhound

We asked 50 top investors what they expect founders to know about their space, then distilled the answers into the Founder Stack. 10 reports backed by ~15 continuous hours of research that go deeper on your market, customers, competitors, wedge, upside, and the case against you than anything else out there. If you don't learn something new, you get your money back.

Get the Founder Stack here: webhound.ai/stacks/founder… $500, and you get $100 of Webhound credit on top to keep researching after.

We asked 50 top investors what they expect founders to know about their space, then distilled the answers into the Founder Stack. 10 reports backed by ~15 continuous hours of research that go deeper on your market, customers, competitors, wedge, upside, and the case against you than anything else out there. If you don't learn something new, you get your money back.

Get the Founder Stack here: webhound.ai/stacks/founder… $500, and you get $100 of Webhound credit on top to keep researching after. That's $600 of value for $500.

Most AI tools hallucinate and hope you don't notice. Webhound now shows the exact tool call chain behind every claim. Every search. Every page. Every python script. Go try to catch it lying. We don't think you will. webhound.ai/news/tool-chai…

Webhound researched a map of internet community reactions to AI. Check it out below webhound.ai/p/ai-community…

@karpathy Would really recommend you try out Webhound, it's essentially what you're describing. webhound.ai

LLM Knowledge Bases Something I'm finding very useful recently: using LLMs to build personal knowledge bases for various topics of research interest. In this way, a large fraction of my recent token throughput is going less into manipulating code, and more into manipulating knowledge (stored as markdown and images). The latest LLMs are quite good at it. So: Data ingest: I index source documents (articles, papers, repos, datasets, images, etc.) into a raw/ directory, then I use an LLM to incrementally "compile" a wiki, which is just a collection of .md files in a directory structure. The wiki includes summaries of all the data in raw/, backlinks, and then it categorizes data into concepts, writes articles for them, and links them all. To convert web articles into .md files I like to use the Obsidian Web Clipper extension, and then I also use a hotkey to download all the related images to local so that my LLM can easily reference them. IDE: I use Obsidian as the IDE "frontend" where I can view the raw data, the the compiled wiki, and the derived visualizations. Important to note that the LLM writes and maintains all of the data of the wiki, I rarely touch it directly. I've played with a few Obsidian plugins to render and view data in other ways (e.g. Marp for slides). Q&A: Where things get interesting is that once your wiki is big enough (e.g. mine on some recent research is ~100 articles and ~400K words), you can ask your LLM agent all kinds of complex questions against the wiki, and it will go off, research the answers, etc. I thought I had to reach for fancy RAG, but the LLM has been pretty good about auto-maintaining index files and brief summaries of all the documents and it reads all the important related data fairly easily at this ~small scale. Output: Instead of getting answers in text/terminal, I like to have it render markdown files for me, or slide shows (Marp format), or matplotlib images, all of which I then view again in Obsidian. You can imagine many other visual output formats depending on the query. Often, I end up "filing" the outputs back into the wiki to enhance it for further queries. So my own explorations and queries always "add up" in the knowledge base. Linting: I've run some LLM "health checks" over the wiki to e.g. find inconsistent data, impute missing data (with web searchers), find interesting connections for new article candidates, etc., to incrementally clean up the wiki and enhance its overall data integrity. The LLMs are quite good at suggesting further questions to ask and look into. Extra tools: I find myself developing additional tools to process the data, e.g. I vibe coded a small and naive search engine over the wiki, which I both use directly (in a web ui), but more often I want to hand it off to an LLM via CLI as a tool for larger queries. Further explorations: As the repo grows, the natural desire is to also think about synthetic data generation + finetuning to have your LLM "know" the data in its weights instead of just context windows. TLDR: raw data from a given number of sources is collected, then compiled by an LLM into a .md wiki, then operated on by various CLIs by the LLM to do Q&A and to incrementally enhance the wiki, and all of it viewable in Obsidian. You rarely ever write or edit the wiki manually, it's the domain of the LLM. I think there is room here for an incredible new product instead of a hacky collection of scripts.

@karpathy Keep an eye out for our next release this week ;)

As a “business guy” I’m slightly embarrassed to share what I’ve been vibe coding. I’ve built a pipeline for doing deep research on San Francisco public policy issues using Manus, Webhound.ai, and Dropbox. When I share I hope you all will have a ton of useful advice (instead of flaming me).

Here's a demo from one of our founders Theo on how we used Webhound Agent to create the Claude skill that one-shotted this video

Introducing Webhound Agent: Your AI research assistant powered by the best deep research money can buy. Most deep research products decide for you when to stop — and since they're subscription-based, the less work they do per query, the more they profit. Webhound flips that. You set the budget and Webhound keeps searching until it's used. $5 or $100 — every dollar goes into deeper search, more sources, and real verification. Today we're launching Webhound Agent — a full research assistant with a workspace. Start research from chat. Analyze results with code and follow-up questions. Organize everything in folders. Run multi-step pipelines. The agent remembers your preferences and gets better the more you use it. Pay-as-you-go. No subscriptions. $5 free to start. See real example reports at different budgets: webhound.ai/examples

A lot of people asking "why webhound?" As one of the founders here's why imo

Webhound's research agent can now write and execute Python code on the fly. Read more here. webhound.ai/news/code-exec…

Introducing cross-session access, conversational planning, and research phases. Read more below. webhound.ai/news/cross-ses…

@frankgoertzen @mwseibel Should be open now

@WebhoundAI @mwseibel DMs closed ?

I always wanted an AI deep research product with the following characteristics: 1. I set the price and it runs for as long as I'm willing to pay 2. Ability to produce reports or structured data 3. Great data validation I think I might have found what I'm looking for: webhound.ai

@frankgoertzen @mwseibel DM the email you used to sign up and we’ll add some

@WebhoundAI @mwseibel Looks fixed. Thanks! Any chance is could get a few credits to re-attempt?

@poorrichard We will look into why this is happening and lyk as soon as it’s fixed. Refunding now.

@WebhoundAI How do I contact your billing team? All I can find is a feedback form and a Discord server. I need help with this.

Very cool stuff!

@poorrichard @frankgoertzen @mwseibel Definitely not normal, but should be fixed now. Thanks for flagging!

@WebhoundAI @frankgoertzen @mwseibel I'm getting... absolutely nothing. Trying to generate a report in guided mode. Got asked one follow-up question right off, and nothing has happened since. No spend (at $0.0065), no plan, nothing. Two hours and counting. Is this normal?